Organ transplantation is now used with great success to improve the quality of human life. Substantial progress has been made in using kidneys, hearts, and livers from unrelated individuals. However, transplantation of hematopoietic stem cells from an unrelated (or allogeneic) donor is a more complicated endeavor. Here multipotent stem cells which have the capacity to regenerate all the blood-forming elements and the immune system are harvested from bone marrow or peripheral blood from one individual are transferred to another. However, histocompatibility differences between donor and recipient results in a higher incidence of transplant-related complications, and has limited the use of this procedure (Forman et al., Blackwell Scientific Publications, 1994).
It is unfortunate that only few individuals are candidates for allogeneic hematopoietic stem cell transplantaiion at the present time because the spectrum of diseases treatable by this procedure has steadily increased. These diseases now include hematologic malignancies such as the acute or chronic leukemias, multiple myeloma, myelodysplastic syndromes; lymphomas; and the severe anemias such as aplastic anemia or thalassemia.
Allogeneic stem cell transplantation begins with treatment of the recipient with a highly immunosuppressive conditioning regimen. This is most commonly accomplished with high doses of chemotherapy and radiation which effectively kill all the blood forming elements of the bone marrow. Besides preparing the recipient bone marrow for donor stem cell transplantation, the conditioning regimen serves to kill much of the malignancy that remains in the body. The period of time between the completion of the conditioning regimen, and engraftment of the donor stem cells is the most dangerous for the recipient. It is during this time that the patient is completely immunocompromised and susceptible to a host of life-threatening infections. This vulnerability persists until the grafted donor stem cells proliferate and differentiate into the needed white blood cells and immune cells needed to combat infections.
Moreover, donor stem cell preparations generally contain immune cells called T lymphocytes. Unless the donor stem cells originate from an identical twin the transferred T cells turn against the recipient's tissues and trigger a deadly ailment called graft versus host disease (or GVHD). This is because the donor T lymphocytes recognize histocompatibility antigens of the recipient as foreign and respond by causing multi-organ dysfunction and destruction.
Current techniques of immunosuppression have made allogeneic stem cell transplantation from a related, histocompatible (HLA-matched) donor much safer than it once was. Allogeneic stem cell transplantation from an unrelated, HLA-matched donor is commonly complicated by serious, often fatal GVHD. The threat of GVHD is even higher when the stem cell donor is HLA mismatched.
Since only 30% of patients in need of allogeneic stem cells will have a sibling with identical histocompatibility antigens (Dupont, B., Immunol Reviews 157:12, 1997), there is a great need to make HLA-matched unrelated, and HLA-mismatched transplantation a safer procedure. There have been two principal approaches to resolving this problem. The first has been to deplete the graft of contaminating T lymphocytes and the second has been to inactivate the T cells so they cannot attack the recipient.
In the 1970's it became evident that ex-vivo removal of mature T lymphocytes from a bone marrow graft prior to transplantation dramatically decreased or prevented GVHD in animals receiving marrow grafts across major histocompatibility barriers (Rodt, H. J. Immunol 4:25-29, 1974; and 4 Vallera et al., Transplantation 31:218-222, 1981). However, with T cell depletion the incidence of graft failure, graft rejection, relapse of leukemia, and viral-induced lympho-proliferative disease markedly increased (Martin et al. Blood 66:664-672, 1985; 6 Patterson et al. Br J Hematol 63:221-230, 1986; Goldman et al. Ann Intern Med 108:806-814, 1988; and Lucas et al. Blood 87:2594-2603, 1996). Thus, the transplantation of donor T cells on the stem cells has beneficial as well as deleterious effects. One needs the facilitating effect of the T cells on the engraftement of stem cells and the now well recognized graft-versus-tumor effects, but not graft-versus host disease.
Several approaches have been used to decrease T cell activation. These include: 1) in vivo immunosuppressive effects of drugs such as FK506 and rapamycin (Blazar et al. J. Immunol 153:1836-1846, 1994; Dupont et al. J. Immunol 144:251-258, 1990; Morris, Ann NY Acad Sci 685:68-72. 1993; and Blazar et al. J Immunol 151:5726-5741, 1993); 2) the in vivo targeting of GVHD-reactive T cells using intact and F(ab′)2 fragments of monoclonal antibodies (mAb) reactive against T cell determinants or mAb linked to toxins (Gratama et al. Am j Kidney Dis 11:149-152, 1984; Hiruma et al. Blood 79:3050-3058, 1992; Anasetti et al. Transplantation 54:844-851, 1992; Martin et al. Bone Marrow Transplant 3:437-444, 1989); 3) inhibition of T cell signaling via either IL-2/cytokine receptor interactions (Herve et al. Blood 76:2639-2640, 1990) or the inhibition of T cell activation through blockade of co-stimulatory or adhesogenic signals (Boussiotis et al. J Exp Med 178:1753-1763, 1993; Gribben et al. Blood 97:4887-4893, 1996; and Blazar et al. Immunol Rev 157:79-90, 1997); 4) the shifting of the balance between acute GVHD-inducing T helper-type 1 T cells to anti-inflammatory T helper-type 2 T cells via the cytokine milieu in which these cells are generated (Krenger et al. Transplantation 58:1251-1257, 1994; Blazar et al. Blood 88:247, 1996, abstract; Krenger et al. J Immunol 153:585-593. 1995; Fowler et al. Blood 84:3540-3549, 1994); 5) the regulation of alloreactive T cell activation by treatment with peptide analogs which affect either T cell receptor/major histocompatibility complex (MHC) interactions, class II MHC/CD4 interactions, or class I MHC/CD8 interactions (Townsend and Korngold (unpublished data)); and 6) the use of gene therapy to halt the attack of donated cells on the recipient's tissues (Bonini et al. Science 276:1719-24, 1997).
There is suggestive evidence that the T lymphocytes from non-identical donors can become tolerant to the recipient's tissues. Unlike patients who receive solid organ allografts for whom life-long immunosuppressive therapy is needed to control chronic rejection, there is evidence of immunologic tolerance with stem cell allografts. The majority of these patients can be withdrawn from immune suppression without further evidence of GVHD (Storb et al. Blood 80:560-561, 1992; and Sullivan et al. Semin Hematol 28:250-259, 1992).
Immunologic tolerance is a specific state of non-responsiveness to an antigen. Immunologic tolerance generally involves more than the absence of an immune response; this state is an adaptive response of the immune system, one meeting the criteria of antigen specificity and memory that are the hallmarks of any immune response. Tolerance develops more easily in fetal and neonatal animals than in adults, suggesting that immature T and B cells are more susceptible to the induction of tolerance. Moreover, studies have suggested that T cells and B cells differ in their susceptibility to tolerance induction. Induction of tolerance, generally, can be by clonal deletion or clonal anergy. In clonal deletion, immature lymphocytes are eliminated during maturation. In clonal anergy, mature lymphocytes present in the peripheral lymphoid organs become functionally inactivated.
Following antigenic challenge stimulation, T cells generally are stimulated to either promote antibody production or cell-mediated immunity. However, they can also be stimulated to inhibit these immune responses instead. T cells with these down-regulatory properties are called “suppressor cells”.
Although it is known that T suppressor cells produce cytokines such as transforming growth factor beta (TGF-beta), interleukin 4 (IL-4) or interleukin (IL-10) with immunosuppressive effects, until recently the mechanisms responsible for the generation of these regulatory cells have been poorly understood. It was generally believed that CD4+ T cells induce CD8− T cells to develop down-regulatory activity and that interleukin 2 (IL-2) produced by CD4+ cells mediates this effect. Although most immunologists agree that IL-2 has an important role in the development of T suppressor cells, whether this cytokine works directly or indirectly is controversial (Via et al. International Immunol 5:565-572, 1993; Fast, J Immunol 149:1510-1515, 1992; Hirohata et al. J Immunol 142:3104-3112, 1989; Taylor, Advances Exp Med Biol 319:125-135, 1992; and Kinter et al., Proc. Natl. Acad. Sci. USA 92:10985-10989, 1995). Recently, IL-2 has been shown to induce CD8+ cells to suppress HIV replication in CD4-T cells by a non-lytic mechanism. This effect is cytokine mediated, but the specific cytokine with this effect has not been identified (Barker et al. J Immunol 156:4476-83, 1996; and Kinter et al. Proc Natl Acad Sci USA 99:10985-9 1995).
A model using human peripheral blood lymphocytes to study T cell/B cell interactions in the absence of other accessory cells has been developed (Hirokawa et al. J. Immunol. 149:1859-1866, 19??). With this model it was found that CD4+ T cells by themselves generally lacked the capacity to induce CD8+ T cells to become potent suppressor cells. The combination of CD8+ T cells and NK cells, however, induced strong suppressive activity (Gray et al. J Exp Med 180:1937-1942, 1994). It was then demonstrated that the contribution of NK cells was to produce TGF-beta in its active form. It was then reported that a small non-immunosuppressive concentration (10-100 pg/ml) of this cytokine served as a co-factor for the generation of strong suppressive effects on IgG and IgM production (Gray et al. J Exp Med 180:1937-1942, 1994). Further, it was demonstrated that NK cells are the principal lymphocyte source of TGF-beta (Gray et al. J Immunol, 160:2248-2254, 1998).
TGF-beta is a multifunctional family of cytokines important in tissue repair, inflammation and immunoregulation (Border et al. J Clin Invest 90:1-7, 1992; and Sporn et al. J Cell Biol 105:1039-1045, 1987). TGF-beta is unlike most other cytokines in that the protein released is biologically inactive and unable to bind to specific receptors (Massague, Cell 69:1067-1070, 1992). The conversion of latent to active TGF-beta is the critical step which determines the biological effects of this cytokine.
There is some evidence that NK cell-derived TGF-beta has a role in the prevention of GVHD. Whereas the transfer of stem cells from one strain of mice to another histocompatibility mismatched strain resulted in death of all recipients from GVHD within 19 days, the simultaneous transfer of NK cells from the donor animals completely prevented this consequence. All the recipient mice survived indefinitely. This therapeutic effect, however, was completely blocked by antagonizing the effects of TGF-beta by the administration of a neutralizing antibody (Murphy et al. Immunol Rev 157:167-176, 1997).
It is very likely, therefore, that the mechanism whereby NK cell-derived TGF-beta prevented GVHD is similar to that described by Horwitz et al. in the down-regulation of antibody production. In each case NK cell-derived TGF-beta was responsible for the generation of suppressor lymphocytes that blocked these respective immune responses. The mouse study is of particular interest since the histocompatibility differences between genetically disparate inbred mice strains would mirror that of unrelated human donors. A modification of this strategy, therefore might overcome GVHD in mismatched humans.
Anti-CD2 monoclonal antibodies and other constructs that bind to the CD2 co-receptor have been shown to be immunosupressive. It has now been demonstrated that at least one mechanism to explain this immunosuppressive effect is by inducing the production of TGF-beta (Gray et al. J Immunol, 160:2248-2254, 1998).
One strategy to prevent GVHD would be to isolate and transfer NK cells along with the stem cells. Another would be to treat the immunocompromised recipient who has received allogeneic stem cells with TGF-beta, anti-CD2 monoclonal antibodies, IL-2 or a combination of these cytokines. The first strategy would be difficult because NK cells comprise only 10 to 20% of total lymphocytes so that it would be difficult to harvest a sufficient number of cells for transfer. The second strategy is limited by the systemic toxic side effects of these monoclonal antibodies and cytokines. IL-2 and TGF-beta have numerous effects on different body tissues and are not very safe to deliver to a patient systemically. What is needed, therefore, is a way to induce mammalian cells to suppress the development of GVHD ex vivo.